Techniques for automatically determining, on a pixel by pixel basis, whether imagery includes ground images or is obscured by cloud cover. The techniques include training a Neural Network, making an initial determination of cloud or ground by using the Neural Network, and performing a max-flow, min-cut operation on the image to determine whether each pixel is a cloud or ground imagery.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
3. The method of claim 2, further comprising connecting the pixels of the grid-graph to at least one of a source or a sink using the score as the capacity, wherein one of the source or the sink represents the cloud imagery and the other of the source or the sink represents the ground imagery, and wherein determining which portions of the image include the cloud imagery and which portions of the image include the ground imagery is further based on the grid-graph.
This invention relates to image segmentation techniques for distinguishing cloud imagery from ground imagery in satellite or aerial images. The problem addressed is accurately separating clouds from the ground in images where both are present, which is challenging due to similarities in texture, color, and lighting conditions. The method involves constructing a grid-graph representation of the image, where each pixel is a node connected to neighboring pixels as edges. A score is computed for each pixel, representing its likelihood of being part of the cloud or ground. These scores are used as capacities when connecting the pixels to a source or a sink in the grid-graph. The source and sink represent the cloud and ground imagery, respectively, or vice versa. By analyzing the flow of data through this graph, the method determines which portions of the image correspond to clouds and which correspond to the ground. The segmentation is refined by leveraging the grid-graph structure, ensuring accurate separation even in complex scenes. This approach improves upon traditional segmentation methods by incorporating spatial relationships and probabilistic scoring to enhance accuracy.
4. The method of claim 1, wherein determining the preliminary classifications is performed using a neural network.
A system and method for classifying data using machine learning techniques addresses the challenge of accurately categorizing large datasets in real-time applications. The invention involves processing input data through a neural network to generate preliminary classifications, which are then refined using additional criteria to produce final classifications. The neural network is trained on labeled data to recognize patterns and features that correlate with specific classifications. The preliminary classifications output by the neural network are evaluated against predefined rules or thresholds to ensure accuracy and consistency. If the preliminary classification meets the criteria, it is accepted as the final classification. If not, the system may trigger further analysis, such as human review or additional processing steps, to resolve discrepancies. This approach improves classification accuracy while maintaining efficiency, particularly in applications where real-time or near-real-time processing is required, such as fraud detection, medical diagnosis, or content moderation. The use of a neural network allows the system to adapt to new data patterns over time, enhancing its performance as more data is processed. The method ensures that classifications are both reliable and scalable, making it suitable for deployment in various industries where automated data classification is critical.
5. The method of claim 1, wherein determining the preliminary classifications is performed by reconstructing each of the multiple portions in a cloud dictionary and in a ground dictionary, and determining the preliminary classifications based on which reconstruction is more accurate.
7. The method of claim 1, wherein the image is a satellite image obtained using a satellite.
8. The method of claim 1, further comprising adding to metadata associated with each pixel an indication of whether each such pixel includes the cloud imagery.
9. The method of claim 8, further comprising selecting, by the processor, pixels for an orthomosaic that is free of clouds based on the metadata.
10. The method of claim 1, wherein determining which portions of the image include the cloud imagery and which portions of the image include the ground imagery is performed by performing a min-cut/max-flow segmentation on the image to define portions of the image that are believed to include the cloud imagery and portions of the image that are believed to include the ground imagery.
14. The method of claim 13, wherein confirming which portions of the image include the cloud imagery and which portions of the image include the ground imagery includes connecting the pixels of the grid-graph to at least one of a source or a sink using the score as the capacity, wherein one of the source or the sink represents the cloud imagery and the other of the source or the sink represents the ground imagery, and wherein confirming which portions of the image include the cloud imagery and which portions of the image include the ground imagery is further based on the grid-graph.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
June 15, 2020
November 15, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.